Steve Yegge's predictions from '04: http://danluu.com/yegge-predictions/. Basically correct on databases, mobile, and cloud computing. Not so much on Lisp

If I look at other tech writing that's held up as well as Steve's 10-20 years later, it's mostly from people who weren't very widely read at the time and are barely read today (e.g., Eric Sink).

Steve was extremely unusual in that he was both influential and correct.

On the flip side, if I look at people who were widely read, when they said things that were falsifiable, they were generally highly confidently but wrong.

This shower thought was inspired by Zach Tellman's recent analysis of one thought leader, but it seems generally true.

For folks who aren't familiar with Eric Sink, here are some old posts:

https://ericsink.com/Abstraction_Pile.html
https://ericsink.com/Pile_Followups.html

He made a bet on C# in early '02. Back then, the preeminent tech thought leader was telling people good programers avoid MS products https://twitter.com/danluu/status/1264018434729238534

Back then, widely read tech bloggers were talking about what makes a Great Hacker (not using NT, for one thing), how you need to understand C pointers, etc

Meanwhile, Eric was talking about how there's more to productivity than being a wizard coder and the importance of teamwork

https://ericsink.com/articles/Choir.html
https://ericsink.com/entries/No_Great_Hackers.html
https://ericsink.com/No_Programmers.html

These posts wouldn't be unusual to see today, but at the time they were written (and for the next decade), when the most widely read tech bloggers were peddling snobbery and elitism, they really stood out.